Works with public HTML pages. Private pages, login-only content, and file downloads are not fetched.
0tokens
0characters
0words
Preview extracted text
The extracted text was limited to 200,000 characters to keep counting responsive.
URL Token Counter
Enter a public webpage URL to extract its readable text and count tokens with the current OpenAI tokenizer.
Paste text
Count private text locally and inspect token boundaries and IDs.
Count files
Measure text, Markdown, JSON, CSV, and code files locally.
Estimate API cost
Turn representative token volume into a monthly model estimate.
FAQ
URL token counter questions
What does the URL token counter include?
It extracts readable text from a public HTML page and removes common navigation, footer, form, script, and style content before counting. Dynamic or login-only text may not be available.
Can it count private or authenticated pages?
No. StackLens does not accept cookies, passwords, local addresses, or private network URLs. Paste private text into the local AI Token Counter instead.
Which tokenizer does the URL counter use?
It uses the current OpenAI text encoding. The result covers extracted plain text, not HTML markup, chat formatting, tools, images, or generated output.
Why can the count differ from copying the page manually?
Webpages can contain hidden, repeated, dynamic, or navigation text. StackLens removes common page chrome and selects the main readable region, so the extracted text may differ from a manual selection.